Transcript ppt - ciera
Population synthesis with dynamics or the problems that we face Natasha Ivanova MODEST-6 August 2005 What is population synthesis? Evolutionary Population Synthesis is the method of direct modeling of large populations of non-interacting objects (single or binaries in our case) with non-trivial (nondescribable by simple analytic) evolution. The evolution for an object is followed from its birth till the desired moment. The goals: a) to see how the population looks like from statistical point of view b) to check if theory works OK and produces the same number of rare objects as observed c) predict new objects! Basics of the binary dynamics Collision time coll: time between two successive collisions, coll = 1/ncS Hardness : ratio between binary binding energy and kinetic energy of an average object Soft binaries <1: get softer (Heggie 1975); very likely to be destroyed through ionization Hard binaries >1: get harder; the encounter can result in the exchange of the companion (smaller mass component is replaced by more massive intruder); very hard binary is likely to merge (Fregeau et al. 2004) Why do we care about binaries? Interacted and destroyed binaries Interacted and survived binaries Non-Interacted binaries Among stars with the initial masses > 0.6 Msun only 8% are both hard and did not have interactions… What do we NEED to take into account for binaries and why life is uncertain Magnetic braking Mass transfer events Mergers Common envelope events Tidal circularization and synchronization Accretion on WDs, Ia SN and subCh Ia SN kicks and NS-retention: e-c SN? Triples?! A few spices before you start to boil the soup IMF Mass ratio distribution? How to pick the secondary? Does q depend on the primary masses? Does q depend on binary periods? What is about “twins”? Periods distribution? Eccentricities distribution? Initial binaries/triples/quadruples/quintuples and sextuples fractions? Some currently existing Population Synthesis codes for interacting binaries Moscow (Yungleson, Tutukov, Lipunov, Postnov,…) Seba (Portegiez Zwart, Tout, Verbunt,…) SSE & BSE (Hurley, Pols, Tout) Kalogera, Belczynski Willems, Kolb … Brussel code (Vanbeveren & Co): interpolation between tracks for massive stars Podsiadlowski, Rappaport, Pfahl & Han : detailed binary tracks for specific classes of systems (Ia SN, LMXBs, sdB …) Any wishes? o fast and robust o correct and powerful (not restricted to a certain class of objects) Time-scale: < one month on a modest 32-CPUs cluster (at 50% Memory-resources: 4Gb of memory (NO swapping!) 250,000 Msun GC: ~ 1,000,000 stars: < 40 seconds per star one star should not take more than 40kb in memory. An unavoidable fact: the evolution of a single star takes much less than the evolution of a binary, if one speaks about an interacting binary. So the code is better be able to evolve a single star on a time-scale of few seconds. GC: evolution is perturbed! SS encounters Mergers in physical collisions Binary formations via physical collisions Binary formations via tidal captures Three and four body encounters MT is interrupted Eccentricity is changed Exchange occurred and companions are now misaligned and not any more on the corotation (multiple) physical collisions, including “dynamical” CE triples formation How to deal with all the mess? Learn binary evolution in the field Build a “scenario processor” for all events Analyze what happens - does it make any sense? Treat specific circumstances with great details Re-run Run to observers Get from observers their results and sit for a while thinking why there is no common language Formation of CV binaries: CE or encounter? Field, non-eccentric binaries CVs formation: main formation channels MS-MS CE BS/BB Destr Exch Merger Collis binary single SS Exch BS/BB CE 40% 10% Coll RG 15% CV 35% CVs: population of WDs Field “Typical” cluster Simulaions vs observations Simulations Observations PS and dynamics: some currently active codes Hurley - open clusters, M67. SeBa - IMBH formation, open clusters ecology Brussel - young clusters with massive stars Freitag - IMBH formation, Galactic center Fregeau, Gurkan, Rasio - IMBH formation, mass segregation in globular clusters with full IMF Ivanova, Belczynski, Fregeau, Rasio - binary fractions, compact binaries formation and evolution in globular clusters Postnov - NS retention, MSPs Future of PS codes with dynamics Bill Paxton’s “EZ” code (Eggleton-refasted): one minute per single star Saul Rappaport and his students at MIT: large population study of binary evolution and RLO. 30,000 binaries in 24 hours at 35 nodes (~2 minutes/RLO binary!!!) http://theory.kitp.ucsb.edu/~paxton